The present application is based on the remarkable finding that a four amino acid extension of a human ribonuclease, eosinophil-derived neurotoxin [EDN(-4)], accounts for much of the cytotoxic activity of crude human chorionic gonadotropin for Kaposi's sarcoma cells . The overall goal of this work is to produce rhEDN(-4) in the chloroplasts of transgenic plants as an abundant source of material for further investigations and clinical development. The rationale for this approach is the finding that previous efforts to express the protein in heterologous systems generate a fully processed hEDN which is devoid of anti-KS cytotoxic activity. Chloroplasts offer a sequestered site, relatively free of signal peptidases, wherein high-level protein expression can be achieved. In phase I we will a) construct chloroplast expression vectors for rhEDN(-4); b) optimize rhEDN(-4) production in tobacco plants; c) evaluate in vitro activity of plant produced rhEDN(-4) and d) evaluate anti-tumor activity of rhEDN(- 4) in a KS-immunodeficient mouse model. Phase II studies of toxicology, GMP production and expanded animal efficacy will be carried out to support filing an IND. rhEDN(-4) may offer a novel, specific, non-toxic therapy for Kaposi's sarcoma and other disorders of angiogenesis. PROPOSED COMMERCIAL APPLICATIONS: The market for an effective agent for Kaposi's sarcoma and other disorders of angiogenesis is substantial.